Important startup metrics to know

In the startup universe, many of us spend hours gazing at the constellation of colorful dashboards and densely arrayed sheets. We obsess over these magical figures, refreshing them incessantly and scanning for clues of movement - a continuous source of joy and misery.

We obsess over these numbers because metrics matter. Good metrics give a snapshot of how well a company is performing and provide signals for course correction.

So what metrics should you actually be looking at? What do VCs look at when analyzing a company? I’ve compiled the key metrics from VCs/investors like Andrew Chen, David Sacks etc, along with my own firsthand experience. Caveat: This list is by no means exhaustive and is likely biased by my own experience in terms of importance. For a broader grasp of startup mental models (covering product to marketing), you can view my previous post here.

Here I’ve split them up into a few broad categories

1. Product/engagement (growth, engagement, retention)
2. Unit economics
3. Financial/business (revenue, margins)
4. Capital efficiency

Product/Engagement

How do you know if your company is hitting the right notes on product or distribution? This first set of metrics is used to determine if your business has reached product market fit (PMF), demonstrated by user growth, engagement, and retention.

1) User growth

For obvious reasons, looking at the current user base and rate of user growth are the first indicators of PMF. It’s the litmus test of market traction. No users = no revenue = end of story.

Despite its ubiquity, it’s still a surprising source of confusion that is usually caused by inconsistent methodologies. So here’s a breakdown of common definitions

• Total users & total customers: Companies tend to conflate these two numbers, which can make it messy internally. Users are typically defined as a successful signup (email or social sign in) or someone that has completed onboarding. In contrast, a customer (or paying/transacting user) is defined as someone that has made some form of transaction. Any “user” definition works, as long as everyone is aligned.

• MAU, DAU: Monthly active users and daily active users. These commonly refer to users that simply are “active” on the app or platform, defined by a unique session. Once again, this “active” definition breeds a lot of confusion, so be super clear on how you define it. For social apps, these numbers are usually the north star metric (and paired with an engagement metric) since total unique eyeballs are part of the game. For e-commerce or other transaction-based apps, MAU is an important but insufficient metric, since they don’t necessarily translate to revenue.

• Activation (or conversion) rate: Signups/paid customer. It’s important to understand if low growth rates are caused by poor distribution (low signups) or a poor onboarding experience (high drop off to paid customer). Signups are a commonly used starting point, but you can go more granular at any step of the funnel (app downloads to signups etc) to diagnose the problem more accurately.

• App downloads: Generally treated as a vanity metric by investors. They will look for actual paying users and more importantly, high engagement. In my experience, it can be a useful immediate signal if something goes off the rails, but the downstream correlation to actual paying users can vary wildly. 

Good growth rates vary depending on the industry and stage. According to YC’s Paul Graham, very early stage (<6 months) startups should be looking at 7% growth WoW - sounds crazy but you need to consider going from 10 to 1000 users. For early stage (1-2 years), it’s not uncommon to do 3x on a yearly basis. As you approach >3 years, growth rates usually go down to around 25% - 100%, which is still considered significant by normal business standards. Stagnant growth rates indicate a lack of PMF or that the distribution strategy is no longer working.

2) Engagement

User growth is just the beginning. There have been many startups that start with exponential growth (Clubhouse as a recent example), but fall off a cliff soon after because users don’t stay engaged with the product. Low engagement leads to lower revenue (if it’s a transaction based business), higher churn, and fewer referrals. Common metrics include

• MAU, DAU: Covered above; a function of both acquisition and engagement.

• Frequency per user: Orders/user (monthly), App opens/user. One of the key indicators used by product and marketing teams as an indicator of how well users are engaging with the product. Typically used to segment audience as well based on activity (active user base: top 10 percentile for orders/month etc)

• DAU/MAU: Underrated but important metric. Calculated by taking the average DAU divided by the MAU on a monthly basis. This ratio paints a good picture of how highly engaged users are. Social platforms tend to have the highest DAU/MAU ratio (20-50%), but you should be comparing against industry averages within your sector or checking your monthly movements. Mixpanel’s product benchmark puts the average app at 12%, which is on the lower end because of the inclusion of SaaS apps - a safer goal should be 15-20%. If you start achieving solid growth and this ratio manages to hold steady, it’s a good sign that newer cohorts are healthy.

Higher engagement > higher propensity to transact > higher revenue + more retention. There are other obvious and important metrics for conversions (monthly orders, transactions, bookings, top-ups etc), and are sometimes used as the north star, but these tend to be very company specific.

3) Churn and retention

Measuring how sticky your users are.

• Monthly churn (Gross): Lost customers/customers from the previous month

• Gross churn vs net churn: Gross churn measures the actual loss of customers, but net churn adds in the new users gained that month. Gross churn should be the main metric because net churn conflates the real picture. Churn can also be looked at from an annual basis, although monthly is more common since a high drop off is a critical red flag that needs to be addressed quickly.

• Retention rate: % of users still active from previous cohort. Retention is simply the inverse of churn, but while churn is usually measured MoM, retention is usually viewed through cohorts. Cohorts are grouped based on signup period (usually by month) and analyzed to see what percentage of users still remain. It’s important to look at cohorts at the 12 to 24 month mark to get a proper sense of retention.

If you’re an early-stage venture, focus on growth first. Activation rates and retention are important and you can start working on small or obvious optimizations. But without sufficient top of funnel traffic and active users, you will not have sufficient sample sizes to test hypotheses and run experiments to improve those metrics meaningfully.

Unit economics

Key financial math for a startup: A look at how much you pay for each customer and how much each customer pays you.

CAC: Customer Acquisition cost

The stuff of every marketer’s nightmare (kidding, sort of..). This is the holy grail metric that anyone within growth or marketing will live and die by. CAC is simply calculated by dividing the total acquisition cost (usually direct marketing expenses + sales cost if any) by the total number of new *paying* customers by the same period.

Blended vs segmented CAC

Deciding what to include in your CAC calculation can be tricky and oftentimes plain misleading, so on top of the total or “blended” (basically the sum of all marketing expenses - brand, perf, etc) CAC, it makes sense to look at acquisition cost or CAC per channel. It’s quite common to show CAC from paid channels separately (total acquisition cost/total customers acquired through paid channels) to evaluate the efficiency of paid channels, especially when organic acquisition is high.

Breaking down CAC by funnel

CAC can also be segmented as CPA based on stages of the funnel (Cost per app install, Cost per signup etc) to get a better sense of where the funnel is breaking. This cost is inversely correlated to the activation rate.

One other note, paid marketing efforts tend to approach diminishing returns with higher spend (there are obvious exceptions to this rule). This means that marginal CAC tends to get higher as you acquire more users. Intuitively this makes sense because the most interested users will first get targeted + organic traction, and as you saturate the audience and scale the budget, you will start having to reach less interested audiences that will cost more to acquire.

LTV: Gross profit per customer (per month) x Average lifespan in months

Lifetime value of a customer is the present cumulative value of all future gross profits from a customer. It’s important to remove any variable costs from the profit here. So if you’re an e-commerce company, use net revenue or margin (GMV x take rate). What form of “profit” can be a source of debate - some VCs suggest that it should be net of all costs, or money that actually flows to the bank.

Average lifespan (months) = 1/Monthly churn 

This number is highly sensitive to your churn input. In early-stage startups or new ventures with less than 2 years of proper customer data, the LTV can vary wildly depending on the denominator used for % churn. My sense is that most people are skeptical about LTVs, simply because they require a multi-year projection and there’s no guarantee that churn rates or ARPU will stay stable over time.

ARPU: Average revenue (usually gross margin) per user

Investors usually use this as a benchmark to see how valuable each user on a platform is. It’s more practical than LTV, because it requires fewer assumptions and can be easily calculated on a monthly basis (simply take the total revenue that month divided by total active users). Sometimes the overall ARPU hides the full picture. A more useful approach is to look at ARPU with a more segmented lens - you can measure the quality of users by looking at the ARPU for each cohort of new monthly users. It’s also a useful tool in understand which customer segments are most lucrative. You can segment your users based on demographic/behavioural attributes and find the correlated ARPU.

There are two main ways to increase ARPU

• Getting more high value clients - this involves redefining a more profitable or relevant target audience and driving a more narrow acquisition strategy. The trade-off here is that a more targeted approach for high quality users will naturally drive CAC up. There is no free lunch.

• Increasing the value from existing users - can be done through marketing initiatives, cross-selling to other platform products,  product enhancements that increase repeated transactions, or simply increasing the price (other obvious trade-offs entail).

Payback period: CAC/ARPU (in months)

Payback period combines ARPU and CAC in a single metric, effectively calculating how many months it will take to earn back what you paid for the customer. In this period where sustainable growth has taken over “growth at all costs”, payback is seen as an important metric to investors. To decrease payback, companies have to navigate a tough line between lowering CAC while improving ARPU. Optimizing for payback usually leads to higher profitability but lower user growth.

Good payback < 6 months
Average < 12 months
Bad < 24 months
Terrible > 36 months

CAC to LTV

The magic ratio that demonstrates the viability of the business. It’s simply the LTV divided by CAC, or a measure of how much a customer is worth vs how much you spent to acquire the customer. The rule of thumb is that anything above 3 is good (ex. $10 CAC and $30 LTV), which means you earn back 3 times based on what you spent to acquire the customer. While this is still a fairly popular metric, it’s fallen out of favour in more recent times because it doesn’t factor in the time it takes to realize that value or the risk of diminishing ARPU.

Business/Financial metrics

This is where the rubber meets the road. The earlier metrics looked specifically at user growth and economics per user, but without real revenue growth and profits at a company level, there isn’t a sustainable business. This usually applies to companies post-seed stage where they need to go beyond user traction and show growth in actual revenues.

Revenue

How much your company earns after actually providing the service or product. Not the same as Bookings (contract value). Generally this is one of the first things that investors look at, followed by the revenue growth trend over the previous periods. For e-commerce the North Star is usually GMV, although there’s a more towards net revenue (gmv x take rate) for a better sense of real profitability.

ARR: Annual recurring revenue

For SaaS companies, ARR and MRR are subscription contract revenues that are recurring. SaaS companies are typically valued with a  multiple of ARR, with median multiples ranging from 5 - 15 depending on industry and annual growth rates. Good sign would be 8 quarters of increasing ARR (Susan Liu)

Gross margins: (Revenue - COGS)/Revenue

Reflected as a percentage and is an indicator of the quality of the revenue. Tech companies generally enjoy good margins because the costs should scale sub-linearly. SaaS companies should have a gross margin of at least 75% (Sacks).

EBITDA: Earnings before interest, taxes, depreciation, and amortization

Used to measure a company’s financial performance, excluding non-operating expenses (financing costs etc) to get a more accurate sense of the company’s operating performance. Most commonly used by investors to benchmark a company against the industry (usually for public comparables). Valuation is typically measured as a multiple of EBITDA.

Oaktree Capital Management’s Howard Marks once said to investors “You can’t eat IRR” in reference to VC distributions. Ultimately in any venture, you can’t eat user growth either, you need revenue.

Capital efficiency

Companies die when they run out of cash. The burn rate tracks the rate that cash is decreasing and is critical for investors to understand the amount of runway a company has left.

• Gross burn: Total monthly cash expenses plus any other cash outlays

• Net burn: Cash revenues minus gross burn

• Runway: Remaining cash/net burn aka how many months you have to live. Most VCs are comfortable with 24 months. Anything below 12 is a red flag.

ROIC: Return on invested capital. Net operating profit/invested capital

Arguably the fundamental metric that’s used to determine how effectively capital is deployed. CAC and LTV can be manipulated based on how you define costs (do you include CS and ops?) and profit (include engineering/legal costs?) but ultimately ROIC tells you how much true profit you get from the amount you spend to get it.

Burn multiple:  Net Burn divided by its Net New ARR in a given period (typically annually or quarterly)

This metric (invented by David Sacks) essentially measures how much the company is spending to generate each dollar of new recurring revenue. Lower burn multiple = more efficient growth. According to Sacks, a Burn Multiple of less than one is amazing, but anything less than two is still quite good. This figure helps to encapsulate more than just sales/marketing costs and looks at how efficient the company is as a whole.

Magic number: Net New ARR/ (Sales + marketing costs)

Widely used by many VCs to evaluate the sales efficiency of SaaS businesses. Ratio should be higher than 1.

Final tips

Having a north star
Ideally, all teams should be aligned to a singular metric depending on the current stage of focus. It can be user growth, GMV, or something company specific like total rides, orders, transaction volume, etc. It’s easy to fall into a trap where more metrics get piled on and everyone is optimizing in different directions. Having a north star doesn’t solve everything, but helps to create focus.

Breaking down metrics
Many high-level metrics are lagging indicators, and might not be useful to guide specific decisions. For individual teams and contributors, it helps to break down the metrics into more granular steps and separate leading/lagging indicators. A large goal like user growth can be broken down into sessions per channel, session length, CTRs, and other conversion rate metrics.

Prevent bad incentives with paired metrics
Sometimes overoptimizing for a certain metric might lead to unintended consequences. Pushing for total new users can lead to poorer quality users. Higher app sessions can lead to lower conversion rates. It’s best to pair a volume metric with a quality one whenever possible. For example, X amount of monthly users while maintaining X ARPU.

Choosing the right metrics for your business, at the right stage
There’s an infamous quote that says: “What gets measured gets managed”. But the full line continues with: " — even when it’s pointless to measure and man- age it, and even if it harms the purpose of the organization to do so". Understand the real business goals and pick the right metrics that are within your control to optimize. If you face astronomical CAC at the early stages, it might mean you haven’t reached product-market fit or that the product requires high trust (therefore larger brand investment) before reaching an inflection point. Optimizing for low CAC right at the beginning could completely stall growth.

Hopefully this provides a good overview of some of the key metrics used by both operators and VCs to gauge the status of the company. I’ll be diving deeper into some metrics with more granularity and tips on how to actually improve them soon.